·onboard.wiki

LLM (Large Language Model)

A Large Language Model is a computer program trained on billions of text documents to predict the next word in a sentence. It doesn’t think, reason, or understand—it recognizes patterns in language well enough to generate text that usually makes sense.

In one sentence

An LLM is pattern-matching tool for text: it reads what you wrote and predicts what words should come next, trained on most of the internet’s public text.

What it's for

LLMs power the chatbots you’ve heard about—ChatGPT, Claude, Gemini. They draft emails, answer questions, summarize articles, write code, and translate languages. They are not general-purpose brains; they are specialized text generators.

How to think about it

Think of an LLM as an extremely well-read intern with no common sense. It has read most of the internet, but it cannot verify facts, lacks real-world experience, and will confidently make things up. Your job is to give it clear instructions and check its work.

Common misunderstandings

Most beginners think LLMs understand what they’re saying. They do not. They also think bigger models are always better—not true for most tasks. A smaller, focused model often outperforms a giant one on specific jobs.

If you want to try it

Open ChatGPT or Claude. Ask it to explain something you know well—your job, a hobby, a local restaurant. Notice where it gets details wrong. That gap between confident-sounding and actually-correct is what you need to watch.

Sources